Southern African Journal o f Gerontology (1992), 1(1): 1-4
Generational interdependence: living arrangements and housing programmes
M. Powell Lawton*
Philadelphia Geriatric Center
Abstract
The broad social issue o f generational competition versus generational interdependence is discussed. The way elders are housed offers an excellent example o f how benefits puta
tively allocated to older people in fact more often than not subsume benefits to fam ily members o f all ages. Data on generationally shared households from a number o f countries and the results o f recent studies from the United States are discussed in this context. Separate housing o f generations is often preferred where feasible. Where economic, environ
mental. health, or social needs o f either elder or young generations make autonomous households dysfunctional, members o f each generation show in their household-forma
tive behaviour their willingness to assist the other generation.
Despite differences in support provisions for older people and differences in the mix between public and private sources of financial support, caring for elders introduces many prob
lems that are very similar in every industrialized country.
Generations and their relationships is a topic that has preoc
cupied gerontologists from the beginning. The classic conclu
sion, demonstrated especially well in the cross-national surveys reported by Shanas et at. (1968), is that modem society has not seen the dissolution of family solidarity.
Although new forms of assistance (for example, caregiving at a geographic distance) may arise, in every country studied by Shanas and her colleagues the overwhelmingly predomi
nant pattern is for the adult children to remain in close touch with aged parents, to offer assistance when needed, and to maintain close affective exchange relationships in the ma
jority of instances.
The conditions under which informal and formal support are delivered to different segments of society may have a major influence on the way generations perceive one another.
The state systems of support for elders have, of course, developed in quite different ways in South Africa, and in the United States. The national government has been a much more active participant in pensions, health, institutional and community-based care in America than in South Africa.
Inevitably, however, increasing industrialization and result
ing social complexity, including increases in costs of social support for disadvantaged groups expected in the future, will raise questions regarding the equitability of public expendi
tures among segmented social groups. Such an issue has been prominent in policy debates in the United States for several years, best known by the somewhat emotion-laden term
"generational equity". Very briefly, a conservative political
movement in the United States is mobilizing pressure to reduce programmes for the aged and their tax support, be
cause other age groups have become disadvantaged in paying for programmes serving the aged. Discussion of this issue may be helpful in anticipating how similar questions might arise and be dealt with in South Africa’s future.
Recent gerontological research has been particularly infor
mative in speaking to questions regarding the distribution of assistance as it goes between generations. The basic question is one of who supports whom, the data leading to the conclu
sion that generational interdependence, rather than gene
rational conflict, characterizes American society and, further, that the same mechanisms probably may be found in South African society. The example of multigenerational dynamics chosen for this article is housing and living arrangements, because of the wide variety of intergenerational exchanges that are possible within the housing context.
Intrafamilial interdependence
National pension systems have transformed patterns of sup
port for older people. In the United States in 1937, two-thirds of the older population was dependent on relatives or public welfare and fully 32 % of all 65+ had no income at all (Upp, 1982). This century saw the advent of Social Security and the growth of employer-administered retirement programmes, as well as federally-assisted health care for elders, all of which changed radically that old style of financial dependence for the older person who no longer worked.
Today few older Americans are dependent in this manner.
In 1978 97 % were “ self-dependent", 1,7 % were dependent on children and 1,3 % on public funds only (Upp, 1982).
In terms of patterns of assistance, people over 65 provide more per capita cash assistance to their children than people under 65 provide for their parents (National Council on the Aging, 1975). A similar situation was reflected in Canadian data from 1978, which showed that the national average annual cash transfer from 65+ parents to an adult child was S213, compared to upward generational transfers to the aged averaging $73 per year (Cheal, 1983). Even older Americans below the poverty level in 1974 (under S3 000) exceeded their children in cash giving (32 % of the poor old gave). The same study showed that a whole series of areas of possible help, such as giving assistance when ill, helping with errands or household tasks, and social-emotional support, are charac
terized by equity and reciprocity of help rather than a unidi
rectional flow from young toward old (National Council on the Aging, 1975).
A d d re ss c o rre sp o n d e n ce to
Prof. M, Powell Lawton, Behavioral Research, Philadelphia Geriatric Center, Old York Road, Philadelphia, PA, United States of America.
1
Southern African Journal o f Gerontology (1992), 1(1)
The area of social relationships is particularly noteworthy.
Contrary to some social stereotypes, the amount of contact between generations is high and continuous: 86 % of older people have living children; 80 % of them live less than an hour’s travel distance from a child; and 70 % had seen one or more children within the past week (Crimmins & Ingegneri,
1990).
Do we need to go any further to illustrate the behavioural togetherness shown by families than these compelling data in the financial and social area? The evidence is clearly on the side of the multigenerational vigour and positive quality of family, while just as clearly those without such resources constitute a target group of high potential need.
Housing and multiple generations
One of the most important areas for intergenerational ex
change is housing and living arrangements.
On the face of it, population data on living arrangements may appear to reflect a dissolution of family solidarity. For example, 30,5 % of Americans 65+ live alone, a proportion that has steadily increased, especially since the 1930s (Sa- luter, 1990). The proportion of older people who lived in the same household with their children was much greater at the turn of this century (Smith, 1981). Furthermore, the usual situation was that the house belonged to the older generation member rather than the younger, an arrangement not always to the liking of the young.
Contrast that situation with the one that holds today. Data that compare different countries in the extent to which older people and their children live together in the same household are informative. In the United States only about 18 % of older people with living children are coresident with their children (Crimmins & Ingegneri, 1990). For contrast, let us consider Japan, where about 67 % live so (Martin, 1989). Data pieced together for other countries indicated that many less indus
trialized countries were similar to Japan in percentage of shared households (China, Korea, Mexico, for example Kin- sella, 1990, and Thailand, South Korea and India, for example Hashimoto, 1991). Southern European countries are only slightly less multigeneration (Greece, Spain, Italy, for example Kinsella, 1990).
Industrialized countries were generally closer to the rate seen in the United States, for example multiple-generation household rates (slightly higher rates as compared to child coresidence rates) were 29 % in France, 27 % in the United
Kingdom and 22 % in Denmark (Kinsella, 1990).
What do we make of these data? First, they seem to support that cultural traditions are an important factor in increasing coresidence, as in the case of Japan. Even more strong, however, is the influence of economic development in de
creasing coresidence. In South Africa a decade ago, the multiple-generation household was as prevalent among blacks as it was anywhere in the world, in contrast to the situation among whites, where the rate was about the same (22 %) as in the United States (Martine, 1979). Today, co
residence is still very prevalent in a rural area of Zimbabwe (76 %) (Hashimoto, 1991). In South Africa however, blacks have become somewhat more involved in the working eco
nomy and considerably freer to relocate, the frequency of multigenerational households has markedly decreased. Even in strongly culturally-determined Japan, the coresidence rate has decreased from 80 % only 20 years ago (Palmore, 1975).
The meaning seems clear; When the state of the country's economy allows it, the generations form separate living units.
It is easy to conclude that American pension and medical benefits have liberated the generations and allowed them to
be able to choose the way they live. As evidenced by their choices, the verdict is clearly toward separateness.
One question worth asking is whether the separateness of the generations is by mutual agreement or by a unilateral decision of the younger family. In general older people are quite happy with their living arrangements. This goes for older people in most circumstances, including, interestingly, those who live alone (Lawton, 1978). Older people, like everyone else, prize their independence, their privacy, and they ding to their own homes. They view their homes as symbols of their lifelong achievement, as objects of attach
ment with which older residents relate themselves in the present to themselves as they were in an earlier period, or as places that are extensions of the Self (Lawton, 1989). Thus at the turn of this century older people typically shared their homes with their children and their grandchildren because there were no alternatives. Today the generations actively enjoy their autonomy. Although the evidence is thus clear that separation of dwellings is by mutual and positive agreement of the majority, there still are forces that tend to produce coresidence.
Two recent studies of different representative American samples have provided us with much-improved under
standing of the dynamics of living arrangements (Aquilino, 1990; Crimmins & Ingegneri, 1990). We have usually as
sumed that the coresidence of an older parent and adult child represented a situation where the elder was in some way dependent upon the adult child. Surprisingly, Aquilino ( 1990) appears to have been the first to characterize whose home it was in which the generations live together. He found that in three-quarters of the instances, the parent was the house
holder. Only one-quarter of the shared households were formed by the child having the parent move into the child’s home, a proportion representing only 4 % of the total elder population. It was in this 4 % of all households containing older people that indicators of major needs of the older person were measurable, needs that could easily be considered served by the younger generation: Widowhood, poverty, and (from Crimmins & Ingegneri’s 1990 data), poor health were very prevalent among these housing-dependent older people.
If parental dependency is not the major determinant of coresidence, what is? Both studies were unanimous in iden
tifying the existence of an unmarried adult child among the living children as the main correlate of there being a shared household. Furthermore, Crimmins and Ingegneri, using data from Shanas’ (1982) survey, reported that more than half of these shared households (56 %) had existed all of the adult child's life. Among the 44 % who had re-formed a shared household after living separately, the reasons for moving back together are shown in Table 1.
With these findings in hand, the reasons for coresidence take on a different look from the view that portrayed the normative situation as one where the adult child provided support for a frail or deprived parent. First, the great prepon
derance of coresidences occur in the parent’s home. Second, the majority of shared households represent states of contin
uous coresidence. Third, the benefits of re-established shared households accrue approximately equally to parent and to child. We lack good information on what caused the contin
uous shared households to remain that way, but it is clear that some segment of this total consists of parents helping deve
lopmentally, mentally, or physically disabled children in their homes and caring for them for a lifetime. Beyond such clear caregiving that flows from the older to the younger generation is a large segment of what we might call coresidence by mutual agreement. The fact that never-married children are strongly overrepresented in this group provides a good 2
Southern African Journal o f Gerontology (1992), 1(1)
example of interdependence that could have resulted from many different mixes of dependence and independence for different families. Presumably some who never married found it easier, less expensive, more comforting, or whatever, to stay in the parental hearth. In other families, a parent may have created a subtle pressure on the child to remain unmar
ried and to stay at home, in the service of the parent’s dependency needs. Still others may simply have wished to stay together and others may have lived under external cir
cumstances that facilitated the maintenance of the status quo with a minimum of active violation on the part of either of both parent and child, In any case, most of the instances of coresidence have been selected by both generations.
Table 1
Reasons for re-formation of shared households
Needs of child Adult child's divorce or widowhood 31 %
Poor health of the child 6%
Other benefits to the child, including economic
15%
52%
Needs of parent Parent's widowhood 17%
Parent's poor health 12%
Other benefits to the parent. 19%
including economic
4 8%
TOTAL 100%
Source: Crimmins & Ingegneri, 1990.
In summary, the ways that people make their living arrangements in the United States show the majority exercis
ing their right to form their own households, where the norm is the nuclear family, the husband-wife pair, or the not pres
ently-married person living alone. Where there are reasons for support being extended, it seems to flow in both directions.
Contrary to both popular and gerontological thought, how
ever, in housing, as well as in cash transfers, the balance is on the side of the housing assistance moving from the elder to the younger family member.
Planned housing and needs of families
The living arrangements discussed so far occur in mainstream housing, i.e. ordinary housing in ordinary communities. The great majority of older people live in such unplanned housing.
Although space will not be taken here to discuss housing planned specifically for elders, this is another important type of housing in the United States. South Africa's planned hous
ing has been developed primarily in the private or private non-profit sector. Nonetheless public support with a form of planning costs, tax loss, and ultimately some use of lax funds, will probably grow in the future. Brief mention of the inter
generational benefits of this type of housing is thus appropri
ate.
Where public costs are involved, one may legitimately ask whether these planned housing units limited to the aged have been a luxury accorded one generation at the expense of the young. The best answer is provided by the answer to another
question. How do younger families wish to live? The answer could n o t b e clearer. “ Intimacy at a distance" (Rosenmayr &
Kockeis, 1963) is a phrase that aptly describes the ideal sociospatial relationship among generations in industrialized countries.
Contrast intimacy at a distance with the alternative. Sharing households is willingly done in response to familial, eco
nomic, emotional, or health-related need but is not the first choice of the majority of people in either generation. In Russia, China and many other countries with major all-gene
ration housing problems one of the major daily irritants is the necessity for young couples to continue to live in a parent's home, often lasting well into the period when the third gene
ration further crowds the tiny flat. Most people in relatively affluent countries have successfully pursued the ideal of privacy and individual space without ever being aware of what the diversity of housing options in their countries' newer housing programmes that include units earmarked for the aged has minimized: Doubting and tripling of children in a single bedroom; a daybed for grandmother in the living room;
the daily kitchen drama of mother and daughter, or, worse yet, daughter-in-law, vying for ascendance; hushed expressions of love, irritation and other emotions between husband and wife. While these are situations with which people cope when they have to, the behaviour of all generations attests that they choose not to do so when possible. For better or for worse, mutual choice has led to geographic separation as the preferred solution to territorial sharing and the risk of conflict.
This scenario characterizes housing for the elderly as a luxury of an affluent society, perhaps, but as a luxury bought for all generations. Advocates fo r the elderly need not ask that housing for the elderly be given greater priority than family housing, but rather that a rational planning process recognize that most often the needs of the generations are served by the same policies.
Conclusion
The answer to possibly different interests of the generations in housing and in other areas lies not in stopping age-specific programmes but in supporting additionally what may have to be quite different initiatives for each segment. Although housing has been used as the example of how preferred mixes of autonomy and support have been achieved by the several generations, similar reasoning could be applied in other sec
tors of life, such as income and health care. To elevate overall quality of life for the poor or disadvantaged younger family the initiative will have to be the dauntingly expensive one of fortifying the roots of the family through education, reward
ing solidarity rather than single parenthood, and putting real money into housing the family, whether by separate gene
ration being a matter of family choice within the bounds of economic reality. Every such successful effort on behalf of the family will have a potential payoff for the older members of such families. A together younger family is in a far better condition to respond to the needs of a grandparent in need than is a fractured problem-ridden single-parent or nuclear family. Just as subsidized housing for the elderly has provided special assistance for vulnerable elders, there is a tremendous need to target the most-vulnerable young - the homeless, the mentally ill, the economically and culturally disadvantaged, the isolated individual - for special assistance.
Effective mobilization of public support for major national investment of effort whether in the public or non-profit sector may be accomplished more readily when support is given by a coalition of all ages, rather than one pitting generation-spe
cific interests against one another.
3
Southern African Journal o f Gerontology (1992), 1(1)
References
Aquilino, W.S. 1990, The likelihood of parent-adult child coresidence:
effects of family structure and parental characteristics. Journal o f Mar
riage and the Family,52: 405-19,
Cheal, D. J . 1983. Intergenerational family transfers. Journal o f Marriage and the Family,45: 805-13.
Crimmins, E.M. & Ingegneri, D.G. 1990. Interaction and living arrangements of older parents and their children. Research on Aging,12: 3-35.
Hashimoto, A. 1991. Living arrangements of the aged in seven developing countries. Journal o f Cross-Cultural Gerontology,6.: 359-81.
Kinsella, K.G. 1990. Living arrangements o f the elderly and social policy: a cross-national perspective,CIR Staff paper No.2, Center for International Research. Washington, DC: Bureau of the Census.
Lawton. M.P. 1978. Housing problems of the community-resident elderly.
Occasional papers in housing and community affairs.No. I . Washington, DC: US Government Printing Office, pp. 39-74.
Lawton, M.P. 1989. Home as an instrument of well-being for older people.
Invited lecture at Canadian Psychological Association annual meeting.
Halifax, Nova Scotia. June.
Martin, L.G. 1989. The graying of Japan. Population Bulletin,44.
Martine, L. 1979. Services for the aged in South Africa. In: Teicher, M.I., Thursz. D. & Vigilante. J.L. (Eds) Reaching the aged: social services in forty-four countries. BeverlyHills, CA: Sage. pp. 227-49
National Council on the Aging. 1975. The myth and the reality o f aging in America.Washington, DC: National Council on the Aging.
Palmore. E. 1975. The status and integration of the aged in Japanese society.
Journal o f Gerontology.30: 199-208.
Rosenmayr. L. & Kockeis, E.1963. Propositions fora sociological theory of aging and the family. International Social Science Journal. 15:410-26 Saluter. A.F. 1990. Marital status and living arrangements. March 1989.
Current Population Reports.Series P-20, No. 445. Washington. DC:
Bureau of the Census.
Shanas. E. 1982. National survey o f the aged. DHSS Pub. No. (OHDS) 83-20425. Washington, DC: Government Printing Office.
Shanas. E.. Townsend, P.. Wedderburn, D.. Friis, H., M ilhoj. P. & Stehouwer, J. 1968. Old people in three industrial societies.New York: Atherton.
Smith, D.S. 1981. Historical change in the household structure o f the elderly in economically developed societies. In: Fogel, R .W , Hatfield, E,. Kiesler.
S.B. & Shanas, E. (Eds) Aging: stability and change in the family.New York: Academic Press, pp. 91-114.
Upp. M. 1982. A look at the economic status of the aged then and now. Social Security Bulletin.45: 16-22.
4
